ACTA UNIVERSITATIS UPSALIENSISAbstracts of Uppsala Dissertations from the Faculty of Medicine

159

MORPHOLOGICAL STUDIES OF THE

VESTIBULAR NERVE

by

b Bjdrn Bergstrim RECEIVEDNASA STI FACIUTY

O p ACQ. BR.

CT 2 9 1973

C 44Q

& c Ob

C'iB

https://ntrs.nasa.gov/search.jsp?R=19730024309 2019-03-25T22:22:42+00:00Z

From the Department of Otolaryngology, University Hospital, Uppsala

MORPHOLOGICAL STUDIES OF THEVESTIBULAR NERVE

Bj6rn Bergstrim

II

MORPHOLOGICAL STUDIES OF THE VESTIBULAR NERVE

CONTENTS

A. INTRODUCTION 7

B. SURVEY OF LITERATURE 9

1. Topographic and gross anatomy of the peripheral vesti-bular system 9

2. The vestibular sensory epithelium 11a. Structure of the sensory cellsb. Morphological polarizationc. Functional polarization and physiological significance

3. The vestibular nerve 15a. The afferent vestibular nerve fibersb. The vestibular ganglion

1. Morphology of the vestibular ganglion2. Tonotopic organization of the vestibular ganglion

c. Central distribution of vestibular nerve fibersd. Efferent fibers to the inner eare. Adrenergic innervation of the inner ear

4. Numerical studies of vestibular nerve fibers, ganglion cellsand sensory elements. Age related changes 22

C. OBSERVATIONS 25I. Anatomical studies of the vestibular nerve in man 25II. The number of myelinated vestibular nerve fibers in man

at various ages 26III. Analysis of the calibers of the myelinated vestibular nerve

fibers in man at various ages 27IV. Macula utriculi and macula sacculi in the squirrel monkey 27

D. DISCUSSION 30

E. SUMMARY 35

F. ACKNOWLEDGEMENTS 37

G. REFERENCES 38

//'

This summary is based on the following papers, which will be referred to bythe Roman numerals I-IV.

I. Bergstr6m, B. 1973. Morphology of the vestibular nerve: I. Anatomicalstudies of the vestibular nerve in man. Acta Otolaryng (Stockh.). In press.

II. Bergstr6m, B. 1973. Morphology of the vestibular nerve: II. Thenumber of myelinated vestibular nerve fibers in man at various ages. ActaOtolaryng (Stockh.). In press.

III. Bergstr6m, B. 1973. Morphology of the vestibular nerve: III. Analysisof the calibers of the myelinated vestibular nerve fibers in man at variousages. Acta Otolaryng (Stockh.). In press.

IV. Engstr6m, H., Bergstr6m, B., and Ades, H.W. 1972. Macula utriculiand macula sacculi in the squirrel monkey. Acta Otolaryng (Stockh.).Suppl. 301, 75-126.

PREEDING PAGESBLAN NOT FILMED

A. INTRODUCTION

During the last decades microsurgery of the temporal bone has come toplay an increasingly important role. Methods and instruments have beendeveloped that make possible operations on branches of the facial andvestibulocochlear nerves and the inner ear without damaging adjacentstructures. The exposure of man and experimental animals to very high orvery low G-forces during certain stages of space flights have also led to anaugmented interest in vestibular function and structure. As a consequencethere is an increasing demand for precise knowledge of the structures in thisarea such as nerves, blood vessels and sensory regions. With these aspectsin mind the following subjects have been included in the present study:

I. A study of the anatomy of the vestibular nerve in man with a survey ofthe literature concerning the peripheral vestibular system. This studyincludes a description of the topographical anatomy of the VIIIth cranialnerve in man with its relations to the inner ear and the facial nerve. It alsocontains a light microscopic study of the vestibular nerve branches and thevestibular ganglion.

II. Against this background it was natural to proceed the project withstudies of possible age related degenerative changes in the vestibular nerve- a controversial question for many years. Since none of the earlier reportsin this field was based on numerical studies it seemed essential to carry out anumerical study of the vestibular nerve and its branches to ascertain:a) what might be termed the "normal" number of nerve fibers andb) whether any reduction in number occurs with increasing age.

III. Based on the observations in part II that there is an age relatedmodification of the vestibular nerve the caliber spectra of the myelinatedvestibular nerve fibers have been analysed in order to determine thedistribution of thick and thin fibers in the ampullary and macular branchesat various ages.

IV. To acquire an understanding of the peripheral course of the vestibularnerve fibers the neuroanatomical studies have been extended down to theultrastructural level. The vestibular sensory cells and their innervation have

7

been the subject of an extensive electron microscopic study which was doneon the squirrel monkey.

The light microscopic and electron microscopic findings in these studieshave been correlated to each other.

8

B. SURVEY OF LITERATURE

1. Topographic and gross anatomy of the peripheral vestibular system

The anatomy of the vestibular nerve has been studied by several authors.Many species of mammals including man have been investigated and foundto have great structural similarities (Retzius, 1884; Voit, 1907; Lorente deN6, 1926; Shute, 1951; Anson et al., 1967; Gacek, 1969; Sando et al.,1972 and others).

The VIIIth cranial nerve (n vestibulocochlearis) contains the afferentfibers from the sensory regions of the inner ear to the vestibular andcochlear nuclei in the brain stem. It also contains efferent fibers to thecochlea (Rasmussen, 1942, 1946) and to the vestibular end organs (Petroff,1955; Rasmussen and Gacek, 1958; Gacek, 1960, 1966) as well asadrenergic fibers (Spoendlin and Lichtensteiger, 1966; Ross, 1971).

According to the Nomina Anatomica (1968) the vestibulocochlear nerveis divided in two main portions, pars vestibularis and pars cochlearis. Parsvestibularis consists of the vestibular nerve except the saccular nervebranch which by the Nomina Anatomica is separated as a special nerve. Noexplanation is given why this branch of the vestibular nerve is considered aseparate nerve. There are no indications in modern literature or in thepresent investigation that the saccular nerve is not a portion of thevestibular nerve.

The vestibulocochlear nerve leaves the brain stem laterally in thecerebello-pontine angle immediately below the lower border of pons.Together with the facial nerve it goes through the internal acoustic porusinto the internal auditory meatus. The nerves are surrounded by a commondural sheath. The facial nerve is positioned antero-superiorly with thevestibulocochlear nerve below and behind (Fig. 1). The fundus region of theinternal meatus is divided in two halves by a bone ridge, the transverse crest.Laterally in the internal auditory meatus the nerves separate; the facial nervecontinues in its own bony canal while the cochlear nerve with its spirallyturned ganglion in the modiolus goes on to the organ of Corti. Thevestibular nerve divides into one superior and one inferior division. Thesuperior division passes with the facial nerve over the transverse crest andthe inferior division runs with the cochlear nerve under the crest. Thevestibular ganglion is situated at the bottom of the internal meatus. Twomain portions of the ganglion can be distinguished, pars superior and

9

nsnap nc

Fig. 1. Schematic drawing of the vestibulocochlear and facial nerves seen from the supero-medial side. "VII" facial nerve. "nc" cochlear nerve. "naa" anterior ampullary nerve. "nal"lateral ampullary nerve. "nap" posterior ampullary nerve. "ns" saccular nerve. "nu"utricular nerve. "nps" superficial petrosal nerves. Arrow indicates vestibulofacial anastomo-sis.

inferior, they are connected by the isthmus ganglionaris (Alexander, 1899).The superior division of the vestibular nerve innervates the sensory epitheliaof the lateral and anterior cristae ampullares and the macula utriculi plus asmall portion of the macula sacculi while the inferior division innervates themain part of the macula sacculi and the posterior ampullary crista.

That a small part of the macula sacculi is innervated by the superiordivision of the vestibular nerve was first observed by Retzius (1881) inteleosts and by Voit (1907) in different species of mammals. The small nerveis usually called Voit's nerve or anastomosis. In mammals including manthere is a connection between the inferior division of the vestibular nerveand the cochlear nerve, the vestibulocochlear anastomosis (Oort, 1918). Itconsists of efferent (Rasmussen, 1946) and afferent (Rasmussen, 1953)

10

fibers to the cochlea. A connection between the lateral ampullary nerve andthe utricular nerve was described by Shute (1951); it was thought tocontribute to the innervation of the macula utriculi. Lorente de N6 (1926)described an anastomosis between the facial nerve and the VIIIth nerve.These fibers were shown to leave the facial nerve at the level of the proximalpart of the geniculate ganglion (Shute, 1951). An anastomosis between thevestibular and facial nerves carries fibers belonging to the intermedius nerve(Gacek and Rasmussen, 1961).

2. The vestibular sensory epitheliuma) Structure ofthe sensory cells

The sensory epithelia on the ampullary cristae and the maculae consist ofhair cells and supporting cells. The hair cells have one long kinocilium andseveral sterocilia arranged in a characteristic pattern. Wersall (1956)described two types of hair cells, type I who were bottle shaped ahdsurrounded by a nerve calyx that almost enclosed the whole cell, and type IIcells that were more cylindrical and were innervated by nerve fibers withbouton shaped nerve endings. The type I cells were mostly seen on top of thecrista while the type II cells were predominantly found in the periphery. Inthe guinea pig Lindeman (1969) found a relation of approximately 3:2between type I and type II cells all over the surface of the crista. Watanukiand Meyer zum Gottesberge (1971) employed different techniques andfound the distribution to be 45 % type I and 55 % type II cells both on top ofthe crista and in the periphery.

On the maculae a lighter zone, the striola (Werner, 1933) can be seeneven at a low magnification in the for the rest homogenous sensoryepithelium. According to Lindeman (1969) the striola is the central, curvedarea where the sensory epithelium and the statoconical membrane have acharacteristic structure (Fig. 2). The hair cells in this area have a muchlarger free surface than the cells outside the striola. Type I cells dominate inthe striola while in the peripheral areas there are about as many type I astype II cells according to Lindeman. Watanuki et al. (1971) also found thattype I cells dominate in the striola but type II cells are so numerous in theperiphery that they outnumber the type I cells on the macula. They found1.2 times more type II than type I cells on the macula sacculi and 1.1. timesmore on the macula utriculi.

The hair cells on the cristae are covered by a gelatinous mass, the cupula,

11

A

C

Fig. 2. Schematic drawings of A) crista ampullaris, B) macula utriculi and C) macula sacculi.

12

into which the hairs penetrate. On the maculae this mass contains agreat number of hexagonal, elongated calcite crystals, statoconiae whichare densely packed in several layers, the statoconial membrane. Sasaki(1970), however, is of the opinion that these crystals are postmortalartifacts and that during life they have a diameter of only 0.1 I.

b) Morphological polarization

The hair cells on the cristae are all orientated in a certain direction, i.e. thekinocilium is localized in the same direction on all cells. The hair cells aresaid to be morphologically polarized in a certain direction. It has been shownthat there is a specific pattern for this polarization. On the crista of thelateral ampulla the cells are polarized towards the utricle while on thecristae of the two vertical ampullae the polarization is directed from theutricle (Lowenstein and Wersall, 1959; Wersill, 1961). On the maculautriculi the hair cells are polarized towards the striola and on the maculasacculi away from the striola (Engstr6m et al., 1962; Flock, 1964; Spoend-lin, 1964, 1965; Lindeman, 1969).

The macula utriculi is almost kidney shaped, the anterior part is oftenslightly broader than the posterior. In the normal anatomical position, i.e. inman the erect position with the head bent 30' forwards the sensoryepithelium lies with its main plane horizontally and the anterior tip slightlyelevated. Thus the hair cells become polarized posteriorly, laterally,medially and in the anterior part, also superiorly and inferiorly.

The macula sacculi is shaped rather like a hook with the anterior part ofthe epithelium bulging outwards in a superior direction. In the normalanatomical position the macula sacculi is orientated almost vertically withthe antero-superior part so situated that the hair cells here face slightlyinferiorly. The polarization will then be mainly in a postero-superior andantero-inferior direction (cf. Lindeman, 1969).

The three semicircular canals are principally positioned in definite planesat right angles to each other. In the normal anatomical position the lateralsemicircular canal lies in the horizontal plane. The ampullary cristae aresaddle shaped with the long axis vertical to the, plane of the semicircularcanal, although minor irregularities have been described (Retzius, 1884;Lindeman, 1969 and others). The sensory cells on the vertical cristae extendfurther down on the utricular than the canalicular side while on the lateralcrista the cells are symmetrically distributed (Wersill, 1956).

13

c) Functionalpolarization and physiological significance

The morphological polarization is accompanied by a functional polarizati-on. An utriculo-petal deviation of the cupula on the lateral crista gives adepolarization of the sensory epithelium (Trincker, 1957) and an increasedimpulse frequency in the nerve fibers (Lowenstein and Sand, 1940). Anutriculo-fugal deviation of the cupula gives a hyperpolarization and adecreased impulse frequency. The opposite conditions are at hand in thevertical cristae.

Gernandt (1949) isolated solitary functional units of the vestibular nervein cat and studied those belonging to the lateral semicircular canal. Hefound 3 different types of response to rotation. In 83 % an increase of theaction potentials was recorded at rotation towards the examined side and adecrease at rotation from the examined side. In 12 % of the cases there wasan increase of the action potentials at rotation in both directions while in5 % a decrease of the action potentials at rotation in both directions wasobserved.

Flock (1964) studied the macula utriculi in a teleost (Lota vulgaris) andconcluded that there is a complicated pattern of interacting afferent andefferent impulses and that the morphological polarization of the hair cellsdecides their directional sensitivity.

Fluur (1970) made experiments with linear accelerations in differentplanes and studied the influence of the maculae on the oculo-motor system.He could also clarify the synergistic and antagonistic interrelationshipbetween different regions of the maculae.

Fluur and Siegborn (1973) have investigated the interaction between theutricles and the horizontal canals. After selective sectioning of the lefthorizontal ampullary nerve they studied how changes in the utricularactivity caused by tilting along the longitudinal axis of the animal influencedthe destruction nystagmus.

Fernandez and coworkers found certain discharge characteristics of theperipheral afferents innervating the cristae and maculae of the squirrelmonkey. They observed a resting discharge which was higher in theampullary nerves (average 90 spikes/sec) than in the macular brancheswhere the utricular nerve had a resting discharge of 79 spikes/sec onaverage and the saccular nerve only 47 spikes/sec. They also found thatmost neurons had a regular discharge pattern and in this group those whohad the highest resting discharge were the most sensitive to stimuli. Inneurons where the discharge was irregular, no such relation was apparent.

14

The irregular neurons included the most sensitive units of all neuronspresent. There were relatively fewer irregular units among the macularneurons than among those to the cristae. The authors suggested that theregular units correspond to the thin fibers innervating the type II cells andthe irregular units correspond to the thick fibers to the type I cells. However,they considered the question unsettled until more unequivocal evidencecould be obtained (Fernandez and Goldberg, 1971; Goldberg andFernandez, 1971 a, 1971 b; Fernandez et al., 1972).

3. The vestibular nervea) The afferent vestibular nerve fibers

The afferent vestibular nerve fibers are myelinated. Lindeman (1969)studied the peripheral distribution of the vestibular nerve fibers. Themyelinated fibers to the cristae show a tendency to run in two branches,oneon the canalicular and one on the utricular side. The utricular nerve fiberspass anteriorly and slightly medially under the sensory epithelium andspread out underneath it. The macula sacculi was found to be innervatedmainly by the saccular nerve but the antero-superior portion was innervatedby Voit's nerve. In the borderline region between the areas of innervation adistinct overlapping of fibers from the two nerve branches was found. Thisoverlapping was said to be most pronounced in man.

Gacek (1969) and Sando et al. (1972) have determined and charteredthe relationships of the afferent vestibular nerve fibers in the internalauditory meatus in the cat. At the distal end of the internal meatus theanterior ampullary nerve lies postero-superior to the lateral ampullarynerve. The utricular nerve is positioned inferior and posterior to theampullary branches. These three branches constitute the superior divisionof the vestibular nerve and they lie above the transverse crest. The saccularnerve and the posterior ampullary nerve form the inferior division where thesaccular nerve lies anterior and superior to the ampullary branch. Theserelationships are principally maintained during the course of the nervefibers to the vestibular ganglion. Central to the ganglion the nerve brancheschange positions and at the level of the porus the posterior ampullary nervelies immediately inferior to the superior ampullary nerves and superior tothe macular branches, the utricular nerve anterior to the saccular.

15

According to Sando et al. (1972) the ampullary nerve fibers rotate in aclockwise direction and the macular fibers in a counterclockwise directionduring their course in the meatus (right ear as seen from the medial side).The cochlear nerve lies inferior and anterior to the vestibular nerve at thetransverse crest and the facial nerve anterior and superior to the vestibularnerve. At the level ofthe porus the vestibular and cochlear nerves have joinedwith the vestibular fibers anterosuperior to the cochlear and the facialnerve lies immediately anterior to the vestibular nerve.

The ampullary branches contain a high proportion of thick fibers. Thethick fibers were found to be related to the top of the cristae and distally theyrun as a central core surrounded by the thinner fibers. The macular nervefibers are mostly thin but a certain amount of thick fibers are scatteredamong these (Gacek, 1969).

b) The vestibular ganglion1. Morphology ofthe vestibular ganglion

In the vestibular ganglion a connection between the ganglion cells of thesuperior and the inferior parts of the vestibular nerve was described byAlexander (1899) who called it the isthmus ganglionaris. Wersill (1956)stated that the vestibular ganglion cells in the guinea pig were localized notonly in the ganglion itself but also were scattered in the isthmus and thenerve branches. According to Ballantyne and Engstr6m (1969), however,the ganglion cells in higher mammals form one ganglion only but a fewganglion cells can occasionally be seen at some distance from the ganglion.They also showed that the vestibular ganglion cells vary considerably insize and that with regard to diameters, there are at least two different popu-lations of myelinated cells. There is one small group with an averagediameter of about 23 i and one large group with an average diameter ofabout 30 a. Apart from the differences in diameter they also found that thecells were of different structure; the majority both large and small weremyelinated but there were also unmyelinated cells. Most cells had a myelinsheath but there was a small number of cells, mostly small that were coveredonly by a simple layer of satellite cell cytoplasm. Those resembled verymuch the cells in the spiral ganglion earlier described by Kellerhals et al.(1967). The cytoplasm is more filamentous than in the myelinated cells andthe endoplasmic reticulum is not so richly developed.

The myelin sheath of the myelinated cells extends along both dendriteand neurite. The appearance of the myelin sheath around the ganglion cell

16

can vary much in different regions and loose lamellae can change withdense even in the same region. The Schwann cell nucleus is often situated ina cytoplasmic out-pouching by the wall of the ganglion cell. The Schwanncells often contain very large mitochondria with densely packed cristae. Theinnermost myelin layers around the vestibular ganglion cell are often soirregular and distorted compared to the nerve fiber sheaths that they have agranular appearance. The myelinated ganglion cell is often ovoid or roundand it has a nucleus in the center or peripherally.

2. Tonotopic organization in the vestibular ganglion

It has been much discussed whether a tonotopic organization existsbetween the sensory cells and the ganglion cells, i.e. that the functionallydifferent areas of the sensory regions have determined projections in thevestibular ganglion as well as in the brain stem nuclei. Lorente de N6 (1926,1931) divided the vestibular ganglion into five regions separated by the sizeof the ganglion cells. These regions were called pars magnocellularisanterior, pars parvicellularis anterior, pars magnocellularis posterior (a),pars magnocellularis posterior (b) and pars parvicellularis posterior. Eachregion was shown to receive fibers from specific parts of the vestibularsensory regions.

Weston (1939) studied the vestibular ganglion in birds, reptiles and fishand could follow the fibers from the cristae to larger ganglion cells thanthose from the maculae.

Werner (1960) found large and small ganglion cells irregularly distribu-ted in the vestibular ganglion in the guinea pig, rabbit and dog.

Gacek (1969) found the cat's vestibular ganglion to form a very compact,linearly arranged mass of cells which spread out in an antero-inferiordirection. Also here a superior and an inferior part could be distinguishedand they were related to the superior and inferior divisions of the vestibularnerve respectively. The ganglion cells related to the macula utriculi occupythe most antero-inferior part of the pars superior while the saccularganglion cells are found in the saccular nerve anterior to the cells for theposterior crista and bordering to the cochlear nerve stem. The ganglion cellsrelated to the anterior and lateral cristae are located partly anteriorly in thesuperior portion of the ganglion and partly postero-inferiorly immediatelysuperior to the ganglion cells for the macula utriculi. Gacek found that thesuperior ampullary nerves receive about the same amount of fibers from

17

both these cell masses and as far as the ganglion is concerned it is one nervewhich peripherally divides into two branches. In the vestibular nerve stemcentral to the vestibular ganglion the fibers from the cristae constitute thesuperior 2/3 of the nerve stem while the macular fibers constitute theinferior 1/3 according to Gacek.

In a study by Stein and Carpenter (1967) the neurites from the superiorvestibular ganglion cells were found to form the superior 2/3 of thevestibular nerve stem with the ampullary fibers running above those fromthe macula utriculi. The inferior 1/3 of the vestibular nerve stem was

composed by the neurites from the inferior ganglion cells with the fibersrelated to the posterior crista positioned superior to those of the maculasacculi.

c) Central distribution of vestibular nerve fibers

After entering the brain stem the neurites of the vestibular nerve divide inascending and descending branches to reach contact with the vestibularnuclei in the bottom of the fourth ventricle. Apart from the four classicalnuclei, nucleus superior, nucleus medialis, nucleus lateralis, and nucleusinferior there is also the interstitial nucleus of the vestibular nerve and anumber of small cell groups called x, y and z.

Within each of the four classical nuclei are regions who do not receiveprimary vestibular fibers but for practical reasons it has been consideredsuitable to retain the accepted nomenclature especially as both "vestibular"and "non-vestibular" parts of the nuclei are very much alike in structureand other fiber connections. In accordance with this reasoning some minorcell groups have been included in the vestibular system even if they do notreceive primary vestibular fibers. Regional cyto-architectonic differences inthe nuclei suggest a functional differentiation which is also evident when thedistribution of fibers to the nuclei is studied and it is also seen inphysiological studies (cf. Brodal et al., 1962).

There is also a central tonotopical organization so that different parts ofthe vestibular sensory regions are projected to separate parts of thevestibular nuclei. Stein and Carpenter (1967) found in their investigation onmonkey that the neurites from those ganglion cells that innervate theampullary cristae are mainly projected to portions of nucleus superior andto the superior parts of nucleus medialis. Central fibers related to theposterior crista were projected to the more medial and inferior parts ofnucleus superior.

18

The neurites from the ganglion cells innervating the macula utriculitraverse, and give collaterals to, anterior parts of nucleus lateralis and reachdown in the postero-medial part of nucleus inferior, collaterals from thesedescending fibers being projected to nucleus medialis. A small number ofthe descending fibers from the utricular part of the superior vestibularganglion pass nucleus inferior and reach nucleus cuneatus accessorius.

The macula sacculi is projected to the postero-lateral parts of nucleusinferior.

Some vestibular ganglion cells are projected to portions of all fourvestibular nuclei so that every special part of the labyrinth has one specificas well as a common projection in the vestibular nuclei. Primary vestibularfibers from all regions of the vestibular ganglion are projected to parts of thenucleus interstitialis.

According to Gacek (1969) the neurites from the ampullary cristae areconnected with the nucleus interstitialis via short collaterals. After theyhave given off these collaterals the neurites continue into the brain stemwhere each axon divides into one ascending and one descending branch.The ascending branches have their nerve endings in nucleus superior andcerebellum while the descending branches run in the vestibular root and giveoff collaterals to nucleus lateralis, medialis and inferior. The neurites fromthe maculae have no connection with nucleus interstitialis or superior. Theascending branches of the utricular neurites have their nerve endings in theantero-superior part of the nucleus lateralis and in the superior part ofnucleus medialis while the descending branches give off collaterals to theinferior part of nucleus medialis and the superior part of nucleus inferior.The nerve endings of the saccular neurites were said to be mainly localizedto the group "y" nucleus with Brodal's nomenclature and to a lesser extentalso to the nucleus lateralis and inferior. Earlier investigations have not beenable to show any afferent fibers to group "y" (Brodal et al., 1962).

The thick fibers from the two superior cristae were found to have theirnerve endings in nucleus superior against the large cells in the center of thenucleus while the thinner fibers terminated in the periphery of the nucleuswhere small cells predominate. Gacek meant that these findings indicatedifferent functional properties for the type I and type II cells since there arereasons to believe that the type I cells are innervated by thick fibers and thetype II cells by thinner fibers. Brodal et al. (1962) found that primaryvestibular fibers also reach the cerebellum where they could be followed toflocculus, nodulus and nucleus fastigii on the homolateral side.

19

d) Efferent fibers to the inner ear

The efferent fibers to the cochlea originate in the contralateral acessorysuperior olivary nucleus (Rasmussen, 1946, 1953) and in the ipsilateralsuperior olivary nucleus (Rasmussen, 1960). The ipsilateral portion passesposterior and lateral to the ascending part of the crossed main portion andunites with it lateral to the root fibers of the facial nerve. Approximately 3/4of the fibers were reported to come from the contralateral side. Rasmussendescribed the efferent fibers to be myelinated, 3-5 pi thick and running asthe olivocochlear bundle in the vestibular nerve. In the cat they were foundto be about 500. However, since then Terayama et al. (1969) and Terayamaand Yamamoto (1971) have shown that the olivocochlear bundle alsocontains unmyelinated fibers in both the crossed and uncrossed portions.The ratio of myelinated and unmyelinated fibers was said to be approxima-

tely 3:2.The olivocochlear bundle leaves the vestibular nerve via the vestibulo-

cochlear anastomosis of Oort and enters Rosenthal's canal where the fibersform intraganglionic spiral bundles. Rossi (1961) demonstrated AChE-positive fibers in the olivocochlear bundle, the intraganglionic spiral bundleand in the organ of Corti. Maw (1973) found terminals of myelinated fibersin the intraganglionic spiral bundle and a branching plexus of unmyelinatedfibers. Wersill (1956) reported two different types of nerve endings to thetype II cells on the cristae, one containing more vesicles than the other. InEngstr6m's (1958) work on the double innervation of the sensory epitheliaof the inner ear two types of nerve endings were shown. One type hadsparsely granulated nerve endings while those of the other type were richlygranulated. It was assumed by Engstr6m that the richly granulated endingswere of an efferent nature and Engstr6m and Fernandez (1961) showed thatafter transection of the crossed efferent bundle to the cochlea a majority ofthe efferent endings disappeared.

Schuknecht et al. (1959) using AChE-technique made degenerationstudies and suggested that the crossed olivocochlear bundle fibers werecontinuous with the unmyelinated fibers and the nerve endings in the organof Corti. It was demonstrated by Wersill et al. (1961) and by Hilding andWersill (1962) that cholinesterase is closely related to the vesiculated nerveendings of cochlear and vestibular hair cells. Evidence of their efferentnature was provided by lurato (1962) and by Kimura and Wersill (1962).

With the AChE-technique Gacek et al. (1965) found a few, 20-30,efferent fibers running in the cochlear nerve trunk. These fibers were traced

20

towards the end organs of the upper middle and apical turns.Petroff (1955) and Rasmussen and Gacek (1958) were the first to

demonstrate an efferent fiber component of the vestibular nerve. Theefferent fibers to the vestibular sensory regions seem to have their origin inthe inferior region of the ipsilateral vestibular nuclear complex (Gacek,1960, 1966). In the vestibular root and nerve these fibers run together withthe efferent cochlear bundle as a compact bundle. Near the vestibularganglion the vestibular efferents divide and run together with and scatteredamong the afferent fibers to the end organs (Gacek et al., 1965). Theyreported the total number of myelinated efferent vestibular fibers to beabout 400 in the cat.

e) A drenergic innervation of the inner ear

The adrenergic innervation of the inner ear has been studied by a number ofauthors. Terayama et al. (1966), Terayama et al. (1968) were of the opinionthat these fibers all came from the ipsilateral superior cervical ganglion andthat they had a perivascular course along the arteries. Spoendlin andLichtensteiger (1966) found apart from this perivascular system, alsoanother and larger adrenergic system which consisted of fibers emerging inthe central nervous system. These fibers reached the periphery via thecochlear nerve with the adrenergic fibers scattered among the afferentcochlear nerve fibers. This latter system is thus independent of the bloodvessels and they reported it to form a peripheral plexus before the habenulaperforata where the afferent fibers begin to become myelinated. In afollowing study Spoendlin and Lichtensteiger (1967) found the perivascularadrenergic system to be a continuous plexus around the vertebral, basilar,inferior anterior cerebellar and labyrinthine arteries reaching as far as themodiolar branches. The second, blood vessel independent system forms arich terminal plexus in the area of the habenula perforata and below thevestibular sensory epithelia. Degeneration studies on cat indicated thatthese fibers originate in the superior cervical ganglion and reach the innerear either via the tympanic plexus-facial nerve-internal auditory meatus orvia the auricular branch of the vagus nerve-facial nerve and internalaudit6ry meatus.

Ross (1971) also found a dual distribution of adrenergic fibers in theinner ear, one group was perivascular and the other went with the afferentcochlear fibers independently of blood vessels. It could not be clearlydecided whether the two groups were of entirely separate origin. In this

21

study was also observed a bundle of adrenergic fibers that reached the innerear via an anastomosis from the carotid plexus to the facial nerve along acourse corresponding to the great superficial petrosal nerve. The bloodvessel independent fibers occurred in greatest numbers within the osseousspiral lamina where some of the fibers formed arcades while othersprojected into the foramina nervosa where they ended as terminal beads.

The functional significance of an adrenergic system in the inner ear hasnot yet been clearly demonstrated. Spoendlin and Lichtensteiger (1966)speculated whether the adrenergic terminal plexus influences the actionpotential of the afferent system either by increasing or decreasing thedischarge threshold. Ross (1971) proposed a generalized function fornorepinephrine in influencing fluid balance and in modifying auditory nerveactivity at the periphery.

4. Numerical studies of vestibular nerve fibers, ganglion cells, and sensoryelements. Age related changes

Numerical studies of the vestibular nerve in man have been done byRasmussen (1940), Bergstr6m (1972) and Naufal and Schuknecht (1972).

Rasmussen counted the vestibular nerve fibers at an intracranial leveland Naufal and Schuknecht sampled the vestibular ganglion cells. Thedistribution of nerve fibers to the different vestibular nerve branches couldnot be determined by either of these methods. Rasmussen divided hismaterial into two groups; one from individuals 2 to 26 years old and thesecond from individuals 44 to 60 years old. In the first group he foundbetween 15.300 and 24.000 vestibular nerve fibers with an average of18.900. In the second group the number of nerve fibers ranged between14.200 and 22.900 with an average of 18.000.

Naufal and Schuknecht found no age related variations in vestibularganglion cell populations. They found an average number of 18.439vestibular ganglion cells (range 14.920-22.390). However, they alsoreported an 86-year-old woman who had only 12.430 ganglion cells in theright and 10.910 in the left vestibular nerve but this reduction was attributedto diabetes mellitus which she had suffered from since the age of 40.

A caliber analysis by Engstr6m and Rexed (1940) showed that thevestibular nerve fibers in general are thicker than the cochlear fibers. 88.5 %of the vestibular nerve fibers had an outer diameter between 2-8 I, 7.2 %were thicker than 8 g and 4.2 % thinner than 2 I while in the cochlear nerve

22

77.1 % of the fibers had an outer diameter between 2-5 t, 14.1 % 5-9 iand 8.7 % were thinner than 2 p. Rasmussen (1940) also found that thecochlear nerve fibers are thinner than the vestibular, in his study thecochlear fibers ranged from 3 to 10 i with the majority 5-7 . Thevestibular fibers varied from 2-15 g, most of them were at least 10 i thick.

The vestibular sensory cells have been analysed by Rosenhall (1972 a,1972 b, 1973). In a group of individuals none of whom was older than 40'years, each of the three cristae had an average number of 7.600 hair cells(range 6.700-8.300). The macula utriculi contained 33.100(29.500-39.200) hair cells and the macula sacculi 18.800(16.000-21.300). In his material from old persons significant reductions innumber of hair cells were found although great individual variationsoccurred. The cristae lost about 40 % of their hair cells on average and themaculae about 20 %.

Apart from Rosenhall's reports the existence of old age related degene-rative changes in the vestibular system has been denied by a number ofauthors. v. Fieandt and Saxen (1937), J6rgensen (1964), Hansen andReske-Nielsen (1964), Reske-Nielsen and Hansen (1964) have all studiedthe degenerative processes of the cochlea in aging. In that connection theyalso looked at the vestibular system without finding any changes that couldbe related to old age.

Schuknecht (1964) presented four cases with extensive degeneration inthe cochlea attributed to old age. In one of the cases he reporteddegeneration of the macula sacculi while the rest of the vestibular systemwas said to be normal. In two of the cases the whole vestibular system wasreported to be normal and in the fourth case there was no information aboutthe vestibular part. Schuknecht et al. (1965) found what they calledcochleo-saccular degeneration in an old cat, an old dog, and an old man. Allthese had extensive age related degeneration in the cochlea and losses ofabout 50 % of the hair cells of the macula sacculi. The other vestibularsensory regions appeared to be normal. They drew the conclusion that thephylogenetically younger pars inferior of the labyrinth is more vulnerable toaging than the phylogenetically older pars superior. Johnsson (1971) andJohnsson and Hawkins (1972 a) found age related degenerations in themacula sacculi but only minor changes in the macula utriculi. However,none of those reports was based on numerical studies. Sercer and Krmpotic(1958) and Krmpotic (1969) have found increasing bone apposition in theforamina for the cochlear and vestibular nerves and the fila olfactoria with

23

increasing age. They are of the opinion that these changes are the primarycause of presbycusis, presbystasis and presbyosmia and the nerve degene-rations are of a secondary character.

24

C OBSERVATIONS

I. Anatomical studies of the vestibular nerve in man

The gross and topographical anatomy of the intratemporal part of the

vestibular nerve in man was studied in 40 temporal bones. After

decalcification the VIIIth and VIIth nerves and the inner ear were dissected.

At the level of the vestibular ganglion the vestibular nerve divides into a

superior and an inferior division. The superior division passes together with

the facial nerve over the transverse crest in the fundus region of the internal

auditory meatus and the inferior division and the cochlear nerve pass under

this crest.

The superior division consists of the lateral and anterior ampullary

nerves and the utricular nerve. The ampullary branches run as a unit for the

greater part of their course and do not divide until just proximal to the

respective ampullae. The utricular nerve is located immediately inferior to

the ampullary branches.

The inferior division contains the saccular nerve and the posterior

ampullary nerve. In one of the investigated nerve specimens the saccular

nerve emitted two branches; one rather thick branch that passed over the

transverse crest to the macula sacculi and one thinner branch which

followed the main saccular nerve under the transverse crest. In practically

all of the investigated nerves the posterior ampulla was innervated by two

branches; the main posterior ampullary nerve and a thin accessory branch

which usually left the vestibular nerve stem a little proximal and superior to

the main branch. The topographical relationships between the vestibulo-

cochlear and facial nerves and the inner ear are described.

A light microscopic study of the vestibular ganglion and the preganglio-

nic part of the vestibular nerve was done on another 12 nerve specimens.

The vestibular ganglion cells form one ganglion divided into a superior and

an inferior portion connected by a narrow isthmus. The superior portion

contains the ganglion cells of the neurons in the superior vestibular nerve

division and the inferior portion those of the inferior division. When

sectioned horizontally the superior portion of the ganglion has a more or

less "Y"-shaped form with a maximum length of about 3 mm. The inferior

portion is more elongated and slender and has a length of 4-4.5 mm. The

isthmus region connects the proximal part of the superior portion with the

central part of the inferior. The ganglion cells have a rich supply of small

blood vessels. The ganglion cells are almost round or ovoid in form with

25

the nucleus more or less in the center. The diameters of the ganglion cellsvary between 15 and 50 li, the majority are 30-40 p in diameter. Regionaldifferences in cell size were not sufficiently evident to allow a subdivision ofthe ganglion based on the size of the ganglion cells.

II. The number of myelinated vestibular nerve fibers inman at various ages

This investigation was performed on the vestibular nerves from 11individuals of varying ages ranging from birth to 85 years. There was noknown history of vestibular disorders, such as Meni6re's disease or acousticneuroma, or of treatment with ototoxic antibiotics or irradiation therapy tothe head.

The aim of the study was twofold; (a) to establish the "normal" numberof nerve fibers in the vestibular nerve and its branches and (b) to investigatewhether any reduction in number occurs with increasing age.

The vestibular nerve fibers occur in bundles irregularly separated byconnective tissue and blood vessels, therefore sampling procedures were notused and all fibers were counted. Control analyses without knowledge of theprevious results differed less than 1% from those.

The material was arranged in three groups according to age. The"young" or "normal" group consists of four individuals, 1 day to 35 yearsold. Two middle-aged persons, 49 and 53 years old constitute the secondgroup. The third or "old age" group consists of five cases, 75-85 years old.

In the young group the number of myelinated vestibular nerve fibersranged from 16.040 to 20.212 with an average of 18.346+3 %. Theseresults agree well with those of earlier reports. The distribution of nervefibers in the different vestibular nerve branches has not been investigatedearlier. In the present study the lateral and anterior ampullary nervestogether contained an average number of 5.899 fibers, the utricular nerve5.952, the saccular nerve 4.046 and the posterior ampullary nerve 2.449fibers. Two thirds of the vestibular nerve fibers are thus found in thesuperior division and one third in the inferior. Although the individualvariations regarding the total number of myelinated vestibular nerve fibersare quite small the variations in the different nerve branches arepronounced.

The two middle-aged cases had 16.582 and 16.817 vestibular nerve fiberswith an average of 16.700.26

In the old age group the number of nerve fibers ranged between 9.274 and15.980 with an average value of 11.506+7 %. This reduction seemed toaffect the ampullary and macular nerve fibers to about the same extent.

The reduction in number of the myelinated vestibular nerve fibers in theold age group averages 37 % when compared to the values of the younggroup. The reduction is statistically significant at better than the 1 % level.

III. Analysis of the calibers of the myelinated vestibularnerve fibers in man at various ages

The caliber spectra of the different vestibular nerve branches were analysed.The material consisted of nerve specimens from one newborn, four adultswith "normal" or near-"normal" numbers of myelinated vestibular nervefibers and four old persons with pronounced reductions in number of nervefibers. These 9 cases were included in study II. The investigation wasperformed on cross sections from the preganglionic parts of the vestibularnerve branches. Photographs were taken through a photomicrosccpe andenlarged to 1000x magnification so that 1 mm on the photographcorresponded to 1 p in the specimen. At this magnification the outerdiameters of the nerve fibers could be measured with accuracy. In all 10.210fibers were analysed. The ampullary nerves and especially the two superiorbranches, were found to contain proportionally more thick fibers (6-15 )than the macular nerves.

In the newborn case the fibers were generally thinner than in the adultand old persons. In the old age group the proportion of thick fibers was lessthan in the adult group, especially in the ampullary nerve branches.

It could not be determined wether these changes were caused bydisappearance of the thick fibers or were the result of a general involution ofall nerve fibers in old age.

IV. Macula utriculi and macula sacculi in the squirrelmonkey

This work contains an ultrastructural study of the vestibular nerve endingsand the vestibular sensory cells. Practical and ethical considerations make itdifficult to obtain human material suited for electron microscopical studies

27

of these regions and therefore the squirrel monkey which has an inner ear

with a high degree of resemblance to the human inner ear (Spoendlin, 1965)was chosen for this study.

The vestibular sensory cells in the more highly developed species

including man are of two types (Wersill, 1956) characterized by their

relations to the afferent nerve endings.The phylogenetically younger type I cell is enclosed in a nerve chalice

which is the terminal portion of a thick myelinated fiber. One nerve fiber can

form separate chalices around a number of type I cells or it can form one

larger chalice to enclose several type I cells. The nerve chalice has a wide

area of contact towards the hair cell.The type II cell receives several club shaped nerve endings which make

synaptic contacts with the cell membrane in many regions mainly in the

infranuclear but also in the supranuclear parts. The afferent nerve endings

are sparsely granulated and contain rather large mitochondria.Many different types of synaptic structures can be seen within the

sensory cells. There are synaptic bars, bodies or balls and also manyintermediate stages between these forms are found.

In the synaptic regions of afferent endings rich invaginations of a "coatedtype" can be observed. In some cases very long synaptic bars, sometimesmore than 5 i long, were observed. In certain regions of type I cells there is

a tendency of a fusion in the synaptic membrane and the possibility ofelectric synapses must be further studied.

The efferent nerve endings are richly granulated and have smallermitochondria than the afferent ones. The type I cells have no direct contactwith the efferent nerve endings which usually are found near the "stem" ofthe nerve chalice. The efferent nerve endings to the type II cells will be foundat many levels from the nucleus and downwards. Usually they are filled withround synaptic vesicles and inside the plasma membrane of the type II cell asubsynaptic cistern is often found.

The terminals of the efferent fibers were seen to run a complicated andrichly branching course in the lower part of the sensory epithelium withgood evidence of "en passant" synapses between efferent fibers both withsensory cells and with afferent fibers. They are readily recognized fromafferent fibers because of the pronounced difference in the size of themitochondria. It is thus directly possible to separate afferent and efferentfibers in the epithelium.

Inside the sensory epithelium the nerve fibers are unmyelinated. Whenthey leave the sensory region a myelin sheath begins within a short distance

28

from the basement membrane. At this level the arrangement of the myelinlayers around the axon with their complicated folds is evident.

Inside the sensory epithelium the afferent fibers to the type I cells areclearly thicker than those to the type II cells.

29

D. DISCUSSION

In the superior division of the vestibular nerve the branches to theanterior and lateral ampullae are almost inseparable during the greater partof their course. In a study on cat Gacek (1969) found that each of the twoampullary nerves was not associated with a distinct portion of the ganglion.He regarded them as one ampullary nerve which splits up peripherally toinnervate two sensory areas. The posterior ampullary nerve was found tohave a thin accessory branch in nearly all of the investigated cases.Montandon et al. (1970) reported two cases of crista neglecta in man wherethey observed an accessory branch to the posterior ampulla. This branchwas said to leave the vestibular nerve stem distal to the main posteriorampullary nerve. They studied more than 600 temporal bones and in all ofthem two posterior ampullary nerves were found. In the present investigati-on the course of the nerve fibers inside the ampulla was not studied and nocrista neglecta was seen.

The saccular nerve can occasionally emit one or two branches which alsoinnervate the macula sacculi. The thick branch observed in one of the nervespecimens in the present material seems to be the same branch thatLindeman (1969) described while the thin branch in the same specimenprobably is the cochleo-saccular nerve described by Hardy (1934) andShute (1951).

The vestibular ganglion cells form one irregularly contoured gangliondivided into a pars superior and a pars inferior connected by a narrowisthmus. A few scattered ganglion cells were found outside the ganglion. Ithas a different form than the spiral ganglion which by Kellerhals et al.(1967) has been compared to a varicose chord. The vestibular ganglion cellsare generally larger than the spiral ganglion cells and show greatervariations regarding cell size (Alexander, 1899; Kellerhals et al., 1967;Ballantyne and Engstr6m, 1969).

It was not possible to make a reliable subdivision of the ganglion intospecific regions based on the cell size as Lorente de N6 (1926) did in the rat.

The nerve specimens in the numerical study (II) were arranged in threeage groups. The young or "normal" group included 4 cases, 1 day to 35years old. This was in accordance with Rosenhall's observations that nosignificant sensory cell degenerations occurred before the age of 40. In thepresent study, however, a certain reduction in number of vestibular nervefibers seems to have taken place already in the 35-year-old case but since

30

individual variations are likely to occur this case has been included in thisgroup and not in the middle-aged group.

The average number of vestibular nerve fibers in this age group agreevery well with those of Rasmussen (1940) and also with Naufal's andSchuknecht's results (1972).

In the old age group a significant reduction in number of nerve fibers wasfound. The average value 11.506 is 37 % less than that of the young group.Individual variations were found; a 75-year-old man thus had 15.980 fibers.The superior and inferior divisions of the vestibular nerve were equallyaffected and so were the saccular and posterior ampullary nerves. Thesuperior division could not be reliably subdivided into ampullary andutricular branches in all cases in the old age group; any possible differencesregarding the degree of reduction in the respective branches could thereforenot be determined. In Rosenhall's studies the hair cells on the cristaesuffered greater losses than those on the maculae.

In the same manner, as has long been known to occur in the acoustic partof the inner ear and VIllth nerve, the vestibular parts undergo pronounceddegenerations with increasing age. It thus seems as if with increasing age,and probably beginning around the age of 40, a general reduction of thenumber of sensory cells and their nerve fibers takes place.

Of great interest is that during the preparation of this paper, van der Laan(1972) of the Jongkees group has shown a reduction of vestibular sensitivitybeginning at the age of 40.

It seems likely that elderly people can suffer from vestibular disordersdue to degeneration of sensory cells and nerve fibers. The fact that vertigoand balance problems are common in old age is probably not caused bythese peripheral lesions only but also by injuries to the central vestibularareas.

Clinical physiological studies have revealed age related changes investibular function (Arslan, 1957; Haas, 1964; Minnigerode et al., 1967;Bruner and Norris, 1971; van der Laan, 1972). Possible explanations of theobserved age effects were thought to be habituation, vascular influences andloss of central inhibition. van der Laan also suggested vestibular nervedegeneration as a possible cause. Vascular changes in the vessels of theinternal auditory meatus and the inner ear with aging have been reported byFisch et al., (1972) and by Johnsson and Hawking, (1972 b).

The ampullary nerves contain a greater proportion of thick fibers thanthe macular branches (III). It was also found that the myelinated fibers ofthe newborn boy were clearly thinner than those of the adult and old

31

individuals. Although the fibers of the 6-week-old boy (II) were not syste-matically measured they also appeared to be thinner than in the older cases.A gradual increase in size of the myelin sheath occurs during a long timeafter birth with increasing age (Westphal, 1897). The myelinization of thevestibular nerve fibers begins at the 20th fetal week and is not completeduntil the time of puberty (Langworthy, 1933). In the old age group adeviation of the caliber spectra towards the thinner side was found,especially in the ampullary nerve branches.

A coarse estimation of the number of sensory cells innervated by eachnerve fiber can be obtained by correlating the hair cell count in the differentvestibular sensory regions with the number of nerve fibers in the nervebranches to these regions.

Lindeman (1969) estimated the number of hair cells in the differentvestibular sensory regions in the guinea pig and Rasmussen and Gacek(1961) have analysed the number of vestibular nerve fibers in the samespecies. Their results are shown and correlated to each other in Tab. I. It isfound that the hair cells on the ampullary cristae are supplied by relativelymore nerve fibers than the macular hair cells, the average ratios being 3.3sensory cells/nerve fiber for the two superior ampullae and 3.6 for theposterior ampulla while on the macula utriculi and macula sacculi thereare 5.4 and 6.0 hair cells to each nerve fiber on average. The same correla-tion between the "normal numbers" of vestibular sensory cells in man re-ported by Rosenhall and the "normal numbers" of vestibular nerve fibersfound in the present study is given in Tab. II. The average ratios are here 2.6hair cells/nerve fiber for the two superior ampullae, 3.1 for the posteriorampulla and 5.6 and 4.6 cells to each nerve fiber for the macula utriculi andsacculi. The hair cells on the ampullary cristae thus seem to be relativelymore richly innervated than those on the maculae in man also.

Table I.Correlation between sensory cells and nerve fibers in the guinea pig.

Guinea Lindeman Gacek/Rasmussen Ratiopig sens. cells nerve fibers cells/fibers

amp ant et lat 11 130 3 389 3,3mac utr 9 260 1703 5,4mac sacc 7 560 1 260 6,0amp post 5 430 1 522 3,6

32

Table II.Correlation between sensory cells and nerve fibers in man.

Man Rosenhall Bergstr6m Ratiosens. cells nerve fibers cells/fibers

amp ant et lat 15 200 5 899 2,6mac utr 33 100 5 952 5,6mac sacc 18 800 4 046 4,6amp post 7 600 2 449 3,1

1

aff e faf

Fig. 3. Afferent (aff) and efferent (eff) innervation of hair cells type I and type II.

33

It is naturally not possible to determine exactly how many hair cells thateach nerve fiber innervates, there are great variations so that one nerve fibermight innervate one or a few hair cells while another fiber is connected withseveral cells as can be seen in electron microscopic studies of the sensoryregions (IV). The type I cell has only one afferent nerve ending but the typeII cell has several. One type II cell can also receive nerve endings fromseveral afferent fibers. It is possible but not probable that one nerve fibergives off a number of nerve endings to a single type II cell. One type I andone type II cell can be innervated by the same nerve fiber and two or moretype I cells can be supplied by a single fiber (Fig. 3).

The current information on the relations between type I and type II cellsin the different vestibular regions is controversial (Lindeman, 1969;Watanuki et al., 1971; Watanuki and Meyer zum Gottesberge, 1971). It istherefore at present not possible to draw any conclusions whether one celltype is relatively more richly innervated than the other.

The sensory cells on the cristae are innervated by proportionally moreand thicker nerve fibers than the maculae. This should be of great functionalimportance and suggests higher demands on the cristae than on themaculae. It is also possible that faster reactions to angular than linearaccelerations are required.

The intraepithelial portions of the fibers to the type I cells are thicker thanthose to the type II cells (IV). If the same caliber relations are present outsidethe sensory epithelium has not been possible to determine in this study.However, if so were the case it would indicate that the cristae containedrelatively more type I cells than the maculae. It would also support the theoriesof Fernandez and coworkers (p. 15) that the regularly firing units correspondto fibers to type II cells and the irregular units correspond to thick fibers totype I cells.

34

E. SUMMARY

I. The gross and topographical anatomy of the intratemporal part of thevestibular nerve in man has been described. In nearly all of the examinedtemporal bones a thin accessory posterior ampullary nerve was seen toleave the vestibular nerve stem a little proximal to and above the mainposterior ampullary nerve. This accessory branch was found to leave theinternal auditory meatus through a separate foramen and it runs for ashorter or longer distance in its own bony canal before uniting with the mainposterior ampullary nerve. In a few cases the two branches coursedseparately all the way to the posterior ampulla.

The form and structure of the vestibular ganglion was studied with thelight microscope. The vestibular ganglion cells form one ganglion dividedinto a superior and an inferior portion connected by a narrow isthmus. Inthe horizontal plane the superior portion is more or less "Y"-shaped whilethe inferior portion has a more elongated and slender form. It was notpossible to subdivide the vestibular ganglion into specific regions based onthe size of the ganglion cells.

The branches of the human vestibular nerve have been sL.c";ed withreference to their general structure.

II. A numerical analysis of the vestibular nerve in individuals of varyingages has been done. In a group of young persons (not older than 35 years)the average number of myelinated vestibular nerve fibers was 18.346. Thehighest values (20.212 and 18.773) were found in a 6-week-old boy and in anewborn boy. The two superior ampullary nerves contained an averagenumber of 5.899 fibers together, the utricular nerve 5.952, the saccularnerve 4.046 and the posterior ampullary nerve 2.449 fibers. The middle-aged cases showed a slight reduction in number of nerve fibers with anaverage value of 16.700 while in the old age group (75-85 years) asignificant reduction down to an average value of only 11.506 fibers wasfound although individual variations occurred. Compared to the younggroup a loss of about 37 % of the myelinated vestibular nerve fibers wasobserved in the old age group.

These findings do not conform with earlier published reports in which agerelated degenerations of the peripheral vestibular system have been deniedapart from the so called cochleo-saccular degeneration. They do, however,conform very well with recently found degenerations of the vestibularsensory regions in old age and also with a recent clinical study. 35

III. The caliber spectra of the myelinated fibers in the vestibular nerve

branches were analysed in nerve specimens from individuals of various ages.The ampullary nerves contain a higher proportion of thick fibers than the

macular nerves. In old age a deviation of the caliber spectra towards the

thinner side occurs.It was also found that the sensory cells on the cristae are relatively more

richly innervated than those on the maculae.

IV. The peripheral endings of the vestibular nerve form a complicated

pattern inside the vestibular sensory epithelia. In an electron microscopicstudy the intraepithelial portions of the afferent and efferent vestibular

nerve fibers have been investigated. The squirrel monkey has been used for

these experiments. As this animal is used to a great extent in our

experimental work a careful study has been made not only of the nerves and

nerve endings but also of the sensory epithelium. This is necessary for a

better understanding of the synaptic morphology. A detailed description is

given of the region of demyelinization and of the intraepithelial course of the

unmyelinated nerve fibers. The granulated or efferent fibers are very distinctfrom the afferent ones. They have much smaller and denser mitochondriaand stand out very clearly (IV, Fig. 35).

The synaptic structures in the squirrel monkey vary considerably andsometimes very conspicious synaptic bars as long as 5 g can be seen. Thereare many different forms of synaptic structures and several types aredemonstrated in this study.

The distance between nerve chalice and type I cell membrane was foundto be in the order of 200-300 A in the non-synaptic regions. In the synapticregions the membranes are considerably closer to each other andsometimes the possibility of a real fusion must be considered. The distancebetween type II cell and afferent nerve endings was 123 A on average in thesynaptic region and the distance between the type II cell and the efferentnerve endings 161 A on average.

The study also contains a detailed description of the sensory cells andtheir surface organelles. An interesting observation is that not only thesensory cells but also many supporting cells are provided with kinocilia. It isalso shown that the cuticular plate is resting on a "shelf'-like reinforcement(IV, Figs. 20, 21, 22). It is further, among other observations, shown thatthe supporting cells contain a fine fibrillar reinforcement reaching from thebasement membrane to the reticular membrane.

36

ACKNOWLEDGEMENTS

The present study was carried out at the Department of Otolaryngology,University Hospital, Uppsala.

I am deeply indebted to my chief and teacher, Professor Hans Engstr6m,head of the Department of Otolaryngology, who provided the original ideafor this work and who has guided me with constructive advice and helpfulsupervision.

I am very grateful to Docent Jan Stahle and Docent G6ran Bredberg,who have offered constructive criticism and valuable advice.

Docent Bj6rn Stenkvist generously put his photomicroscopic equipmentat my disposal.

The invaluable help and advice given by Berit Engstr6m, ResearchEngineer, in technical problems, photography and schematic drawings isgratefully acknowledged.

I am grateful to Docent Bo Jung who helped me with statistical problems.I also wish to express my gratitude to Mrs. Gunilla Rajamiki and Mrs.

Lena Svedjebick for their technical assistance; to Mrs. Ulla Wikberg andMrs. Marianne Jonsson who have typed the manuscripts.

I also wish to thank Dr. Richard Maw and Mr. Ivan Hunter-Duvar,Ph.D., who have supervised and corrected the English text. Miss HelgaM6bius is acknowledged for the translation of the "Zusammenfassungen".

This study has been supported in part by NASA Grants No. NCR-52-028-003 and NGR 52-028-004 and by the University of Uppsala.

37

REFERENCES

Alexander, G. 1899. Zur Anatomie des Ganglion Vestibulare. Sitzungsberichte d kais Akad d Wiss,

math-naturw Classe 108, 449.

Anson, B.J., Harper, D.G. & Winch, T.R. 1967. The vestibular system. Anatomic considerations. Arch

Otolaryng 85, 497.

Arslan, M. 1957. The senescence of the vestibular apparatus. Pract Otorhinolaryng (Basel) 19, 475.

Ballantyne, J. & Engstr6m, H. 1969. Morphology of the vestibular ganglion cells. JLaryng 83, 19.

Bergstr6m, B. 1972. Numerical analysis of the vestibular nerve in man. A preliminary report. UpsalJ

Med Sci 77, 205.

Brodal, A., Pompeiano, O. & Walberg, F. 1962. The vestibular nuclei and their connections, anatomy

and functional correlations. The Henderson Trust Lectures. Edinburgh and London: Oliver and

Boyd.

Bruner, A. & Norris, T.W. 1971. Age-related changes in caloric nystagmus. Acta Otolaryng (Stockh.)

Suppl. 282, 1.

Engstr6m, H. 1958. On the double innervation of the sensory epithelia of the inner ear. Acta Otolaryng

(Stockh.) 49, 109.

Engstr6m, H. & Rexed, B. 1940. Uber die Kaliberverhiltnisse der Nervenfasern im N. stato-acusticus

des Menschen. Z mikrosk-anat Forsch Akad Verlagsgesellschaft. Leipzig 47, 448.

EngstrBm, H. & Fernandez, C. 1961. Discussion to Catherine A. Smith. Trans Amer Otol Soc 49, 58.

Engstr6m, H., Ades, H.W. & Hawkins, J.E. Jr. 1962. Structure and functions of the sensory hairs of the

inner ear. J Acoust Soc Amer 34, 1356.

Fernandez, C. & Goldberg, J.M. 1971. Physiology of peripheral neurons innervating semicircular

canals of the squirrel monkey. II. Response to sinusoidal stimulation and dynamics of peripheral

vestibular system. J Neurophysiol 34, 661.

Fernandez, C., Goldberg, J.M. & Abend, W.K. 1972. Response to static tilts of peripheral neurons

innervating otolith organs of the squirrel monkey. J Neurophysiol 35, 978.

v Fieandt, H. & Saxen, A. 1937. Pathologie und Klinik der Altersschwerh6rigkeit. Acta Otolaryng

(Stockh.) Suppl. 23, 1.

Fisch, U., Dobozi, M. & Greig, D. 1972. Degenerative changes of the arterial vessels of the internal

auditory meatus during the process of aging. Acta Otolaryng (Stockh.) 73, 259.

Flock, A. 1964. Structure and function of the macula utriculi with special reference to directional

interplay of sensory responses as revealed by morphological polarization. J Cell Biol 22, 413.

Fluur, E. 1970. The interaction between the utricle and the saccule. Acta Otolaryng(Stockh.) 69, 17.

Fluur, E. & Siegborn, J. 1973. Interaction between the utricles and the horizontal semicircular canals.

Acta Otolaryng (Stockh.) 75, 17.

Gacek, R.R. 1960. Efferent component of the vestibular nerve. In: Neular mechanisms of the auditory

and vestibular systems. (G.L. Rasmussen and W.F. Windle, eds.) Charles C. Thomas, Springfield,

276.- 1966. The vestibular efferent pathways. In: The Vestibular System and Its Diseases. (R.J. Wolfson,

ed.) University of Pennsylvania Press.

- 1969. The course and central termination of first order neurons supplying vestibular end organs in

the cat. Acta Otolaryng (Stockh.) Suppl. 254, I.

Gracek, R.R. & Rasmussen, G.L. 1961. Fiber analysis of the statoacoustic nerve of guinea pig, cat, and

monkey. Anat Rec 139, 455.

38

Gacek, R.R., Nomura, Y. & Balogh, K. 1965. Acetylcholinesterase activity in the efferent fibers of the

stato-acoustic nerve. Acta Otolaryng (Stockh.) 59, 541.

Gernandt, B. 1949. Response of mammalian vestibular neurons to horizontal rotation and caloric

stimulation. J Neurophysioll12, 173.

Goldberg, J.M. & Fernandez, C. 1971 a. Physiology of peripheral neurons innervating semicircular

canals of the squirrel monkey. 1. Resting discharge and response to constant angular accelerations.

J Neurophysiol 34, 635.

- 1971 b. Physiology of peripheral neurons innervating semicircular canals of the squirrel monkey.

III. Variations among units in their discharge properties. J Neurophysiol 34, 676.

Haas, E. 1964. Zur Frage der Altersabhlngigkeit der Drehreizschwelle. Z Laryng Rhino143, 238.

Hansen, C.C. & Reske-Nielsen, E. 1964. Pathological studies in perceptive deafness. A patient with

hydrops of the labyrinth. Acta Otolaryng (Stockh.) Suppl. 188, 162.

Hardy, M. 1934. Observations on the innervation of the macula sacculi in man. A nat Rec 59, 403.

Hilding, D. & Wersill, J. 1962. Cholinesterase and its relation to the nerve endings in the inner ear. Acta

Otolaryng (Stockh.) 55, 205.

Iurato, S. 1962. Efferent fibers of the sensory cells of Corti's organ. Exp CellRes 27, 62.

Johnsson, L.-G. 1971. Degenerative changes and anomalies of the vestibular system in man.

Laryngoscope 81, 1682.

Johnsson, L.-G. & Hawkins, J.E. Jr. 1972 a. Sensory and neural degeneration with aging, as seen in

microdissection of the human inner ear. Ann Otol 81, 179.

- 1972 b. Vascular changes in the inner ear associated with aging. Ann Otol81, 364.

J6rgensen, M.B. 1964. Changes of aging in the inner ear, and inner ear in diabetes mellitus. Histological

studies. Acta Otolaryng (Stockh.) Suppl. 188, 125.

Kellerhals, B., Engstr6m, H. & Ades, H.W. 1967. Die Morphologie des Ganglion Spirale Cochleae.

Acta Otolaryng (Stockh.) Suppl. 226, 1.

Kimura, R. & Wersill, J. 1962. Termination of the olivocochlear bundle in relation to the outer hair cells

of the organ of Corti in guinea pig. Acta Otolaryng (Stockh.) 55, 11.

Krmpotic, J. Nemanic. 1969. Presbycusis, presbystasis and presbyosmia as consequences of the

analogous biological process. Acta Otolaryng (Stockh.) 67, 217.

van der Laan, F.L. 1972. Age and vestibularfunction. Boek-en Offsetdrukkerij Mur-Aalsmeer.

Langworthy, O.R. 1933. Development of behavior patterns and myelinization of the nervous system in

the human fetus and infant. Contr Embryol 24, 1.

Lindeman, H.H. 1969. Studies on the morphology of the sensory regions of the vestibular apparatus.

Ergebn der Anatomie 42, 1.

Lorente de N6, R. 1926. Etudes sur I'anatomie et la physiologie du labyrinthe de l'oreille et du VIIIe

nerf. Deuxieme partie. Quelques donn6es au sujet de l'anatomie des organes sensoriels du

labyrinthe. Trav Lab Rech Biol Univ Madrid 24, 53.

- 1931. Ausgewihlte Kapitel aus der vergleichenden Physiologie des Labyrinthes. Die Augenmuskel-

reflexe beim Kaninchen und ihre Grundlagen. Ergebn Physiol 32, 73.

Lowenstein, O. & Sand, A. 1940. The mechanism of the semicircular canal. A study of the responses of

single fiber preparations to angular accelerations and to rotation at constant speed. ProcRoySoc B

129, 256.

Lowenstein, O. & Wersill, J. 1959. A functional interpretation of the electron-microscopic structure ofthe sensory hairs in the cristae of the elasmobranch Raja clavata in terms of directional sensitivity.

Nature (Lond.) 184, 1807.

39

Maw, R. 1973. Further morphological modifications of the concept of efferent cochlear innervation at

an ultrastructural level. J Laryng (in press).

Minnigerode, B., Grohmann, R. & Vontin, H. 1967. Elektronystagmographische Untersuchungen zum

Verhalten der Labyrinthfunktion gesunder Versuchspersonen verschiedenen Lebensalters. Pract

Otorhinolaryng (Basel) 29, 153.

Montandon, P., Gacek, R.R. & Kimura, R.S. 1970. Crista neglecta in the cat and human. Ann Otol 79,

105.

Naufal, P.M. & Schuknecht, H.F. 1972. Vestibular, facial, and oculomotor neuropathy in diabetes

mellitus. Arch Otolaryng 96, 468.

Nomina Anatomica. 1968, 3rd edition. Excerpta Medica Foundation.

Oort, H. 1918-19. Uber die Veristelung des Nervus octavus bei Siugetieren. A nat A nz 51, 272.

Petroff, A. 1955. An experimental investigation of efferent fiber projections to the vestibular

neuroepithelium. (Abstract) Anat Rec 121, 352.

Rasmussen, A.T. 1940. Studies of the VIIlth cranial nerve of man. Laryngoscope 50, 67.

Rasmussen, G.L. 1942. An efferent cochlear bundle. (Abstract)AnatRec82, 441.

- 1946. The olivary peduncle and other fiber projections of the superior olivary complex. J comp

Neurol 84, 141.

- 1953. Further observations of the efferent cochlear bundle. J comp Neuro1l99, 61.

- 1960. Efferent fibers of the cochlear nerve and cochlear nucleus. In: Neural Mechanisms of the

Auditory and VestibularSystems. (eds. G.L. Rasmussen and W.F. Windle), Springfield I 11, 105.

Rasmussen, G.L. & Gacek, R.R. 1958. Concerning the question of an efferent fiber component of the

vestibular nerve of the cat. (Abstract) Anat Rec 130, 361.

Reske-Nielsen, E. & Hansen, C.C. 1964. Pathological studies in perceptive deafness. Four patients with

presbyacusis. Acta Otolaryng (Stockh.) Suppl. 188, 327.

Retizius, G. 1881. Das Gehdrorgan der Wirbelthiere. I Das Gehdrorgan der Fische und Amphibien.

Der Centraldruckerei, Stockholm.

- 1884. Das Gehororgan der Wirbelthiere. II Das Gehororgan der Reptilien, der Vogel und der

Sdugethiere. Der Centraldruckerei, Stockholm.

Rosenhall, U. 1972 a. Vestibular macular mapping in man. Ann Otol81, 339.

1972 b. Mapping of the cristae ampullares in man. Ann Otol81, 882.

- 1973. Degenerative patterns in the aging human vestibular apparatus. In manuscript.

Ross, M.D. 1971. Fluorescence and electron microscopic observations of the general, visceral, efferent

innervation of the inner ear. Acta Otolaryng (Stockh.) Suppl. 286, 1.

Rossi, G. 1961. L'acetylcholinesterase au cours du developpement de l'oreille interne du cobaye. Acta

Otolaryng (Stockh.) Suppl. 170, 1.

Sando, I., Black, F.O. & Hemenway, W.G. 1972. Spatial distribution of vestibular nerve in internal

auditory canal. Ann Otol 81, 305.

Sasaki, H. 1970. Basic and clinical studies of the otolithic organ. JOtolaryngJap 73 Suppl 1278.

Schuknecht, H.F. 1964. Further observations on the pathology of presbycusis. Arch Otolaryng 80, 369.

Schuknecht, H.F., Churchill, J.A. & Doran, R. 1959. The localization of acetylcholinesterase in the

cochlea. Arch Otolaryng 69, 549.

Schuknecht, H.F., Igarashi, M. & Gacek, R.R. 1965. The pathological types of cochleosaccular

degeneration. Acta Otolaryng (Stockh.) 59, 154.

Sercer, A. & Krmpotic, J. 1958. Uber die Ursache der progressiven Altersschwerhrigkeit

(presbycusis). Acta Otolaryng (Stockh.) Suppl. 143, 1.

40

Shute, C.C.D. 195 1. The anatomy of the eighth cranial nerve in man. Proc Roy Soc Med 44, 1013.Spoendlin, H. 1964. Organization of the sensory hairs in the gravity receptors in utricle and saccule of

the squirrel monkey. Z Zellforsch 62, 701.- 1965. Ultrastructural studies of the labyrinth in squirrel monkeys. In: The Role of the Vestibular

Organs in the Exploration of Space. NASA SP-77, 7.Spoendlin, H.H. & Lichtensteiger, W. 1966. The adrenergic innervation of the labyrinth. Acta

Otolaryng (Stockh.) 61, 423.- 1967. The sympathetic nerve supply to the inner ear. Arch klin exp Ohr-Nas-Kehlk Heilk 189, 346.Stein, B. & Carpenter, M. 1967. Central projections of portions of the vestibular ganglia innervating

specific parts of the labyrinth in the Rhesus monkey. Am JAnat 120, 281.Terayama, Y. & Yamamoto, K. 1971. Olivo-cochlear bundle in the guinea pig after central transsection

of the crossed bundle. Acta Otolaryng (Stockh.) 72, 385.Terayama, Y., Holz, E. & Beck, C. 1966. Adrenergic innervation of the cochlea. Ann Otol75, 69.Terayama, Y., Yamamoto, K. & Sakamoto, T. 1968. Electron microscopic observations on the

postganglionic sympathetic fibers in the guinea pig cochlea. Ann Otol 77, 1152.- 1969. The efferent olivo-cochlear bundle in the guinea pig cochlea.Ann Otol 78, 1254.Trincker, D. 1957. Bestandspotentiale im Bogengangssystem des Meerschweinchens und ihre

Anderungen bei experimentellen Cupula-Ablenkungen. Pfligers Arch ges Physiol 264, 351.Volt, M. 1907. Zur Frage der Verdstelung des Nervus Acusticus bei den Siugetieren. A natAnz 31, 635.Watanuki, K. & Meyer zum Gottesberge, A. 1971. Light microscopic observations of the sensory

epithelium of the crista ampullaris in the guniea pig. Ann Otol 80, 450.Watanuki, K., Stupp, H.F. & Meyer zum Gottesberge, A. 1971. Distribution pattern of the type I and

type II sensory cells on the maculae sacculi and utriculi in the guinea pig. Pract oto-rhino-laryng 33,304.

Werner, C.F. 1933. Die Differenzierung der Maculae im Labyrinth, inbesondere bei Siugetieren. ZAnat Entwicklungsgesch 99, 696.

- 1960. Das Gehdrorgan der Wirbeltiere und des Menschen. Thieme, Leipzig.Wersill, J. 1956. Studies on the structure and innervation of the sensory epithelium of the cristae

ampullares in the guinea pig as revealed by the electron microscope. Acta Otolaryng (Stockh.)Suppl. 126, 1.

- 1961. Vestibular receptor cells in fish and mammals. Acta Otolaryng(Stockh.) Suppl. 163, 25.Wersill, J., Hilding, D. & Lundquist, P.-G. 1961. Ultrastruktur und Innvervation der cochlearen

Haarzellen. Arch klin exp Ohr-Nas-Kehlk Heilk 178, 106.Weston, J.K. 1939. Observations on the comparative anatomy of the VIIIth nerve complex. Acta

Otolaryng (Stockh.) 27, 457.

Westphal, A. 1897. Uber die Markscheidenbildung der Gehirnnervendes Menschen. Arch Psychiat 29,474.

41